With the rapid development of computer technology and automation technology, conducting automation tests through scripts is a currently popular test approach. Typically a test object map is always associated with a script in an automation test tool. When a tester catches an object by an object finder, the selected object and its hierarchy (including siblings) can be saved in the test object map. However, since a test cycle of the script is usually long and objects in the script might change in the development phase (for example, since research and development staff members revise the script), some objects cannot be recognized any more (missed). Thus, it happens that testers run scripts for a long time but find that most of them fail because individual objects cannot be recognized any more. The failed object recognition blocks all following steps.
In the prior art, there are three common solutions to the above problem: letting research and development staff members inform testers every time and then revise scripts; upon test failure, making comparison by manually catching using tools such as QTP provided by Hewlett-Packard and the like, so as to learn the reasons for the failure; and manually debugging scripts. Obviously these solutions are laborious and time-consuming, which greatly decreases the test efficiency.